FACTORS FOR YIELD AND QUALITY OF SPRING WHEAT 7 
dwarfs and grass clumps, the two latter not being distinguishable. 
These results indicate that two genetic factors are involved, segre- 
gating in the F 2 progenies to give a 13 : 3 ratio for normals to com- 
bined dwarfs and grass clumps and in a 15 : 1 ratio for combined 
normals and dwarfs to grass clumps. As the Moccasin data do not 
permit the latter separation, the F 2 data from both stations are 
tabulated in only two classes — 'normals compared with dwarfs, the 
latter including grass clumps. The data for each of the Fj families, 
the totals at each station, and the totals from both stations are given 
in Table 3. 
Table 3. — Segregation of 1,310 F 2 plants of the Marquis X Hard Federation 
wheats and reciprocal crosses into two classes, normal and dwarf plants, when 
grown at Bozeman and Moccasin, Mont., in 1923 
Number 
of Fi 
families 
Fj plants 
Devia- 
tion 
from 
13:3 
ratio 
Probable 
error 
Source, cross, and hybrid No. 
Normals 
Dwarfs 
Total 
BOZEMAN 
MarquisXHard Federation: 
21202B1 
1 
96 
100 
101 
27 
28 
26 
123 
128 
127 
4 
4 
2 
2.92 
21202B2 
2.98 
21202B3 
2.97 
Total: 
Number 
Percentage 
3 
297 
78.6 
81 
21.4 
378 
10 
5.12 
Hard FederationXMarquis: 
21203A1 
1 
1 
1 
102 
101 
100 
24 
27 
20 
126 
128 
120 

3 
3 
2.96 
21203A2 ._ 
2.98 
21203A3 
2.88 
Total: 
3 
303 
81.0 
71 
19.0 
374 
1 
5.09 
Percentage 
Grand total: 
6 
600 
79.8 
152 
20.2 
752 
11 
7.22 
Percentage .... . . 
MOCCASIN 
MarquisXHard Federation: 
21202A1 
1 
1 
1 
81 
72 
77 
10 
22 
18 
91 
94 
95 
7 
4 

2.51 
21202A2 
21202A3 
2.55 
2.57 
Total: 
Number 
Percentage 
3 
230 
82.1 
50 
17.9 
280 
3 
4.41 
Hard FederationXMarquis: 
21203C1 
21203C2 
21203C3 
1 
81 
70 
76 
15 
15 
21 
96 
85 
97 
3 
1 
3 
2.58 
2.43 
2.59 
Total: 
Number 
Percentage ... 
3 
227 
81.7 
51 
18.3 
278 
1 
_ 
4.39 
Grand total: 
Number 
Percentage .. 
■ 
6 
457 
81.9 
101 
18.1 
558 
4 
6.22 
Total (both stations) : 
Number 
Percentage. 
12 
1,057 
80.7 
253 
19.3 
1,310 
7 
9.53 
The data show a close fit to the 13 : 3 ratio in each F x family and 
in their totals at both stations. The deviations in numbers in all 
cases are less than three times their probable errors. There is no 
